Study population and sample collection.
The study population was recruited among persons seeking care at Mercy Hospital in Bo, Sierra Leone. All persons with clinically confirmed or self-reported fever with onset within the 10 days before the enrollment date were invited to participate in the study. Informed consent from patients (or, for children, consent from their parents) was obtained and documented prior to collection of clinical data and biological specimens. In total, 141 volunteers were enrolled between 28 January and 20 May 2019. Paired samples of venous and capillary blood were collected into EDTA-containing vacutainers or microtainers from each participant by venipuncture and finger prick, respectively.
The study protocol was approved by the Institutional Review Boards at the US Naval Research Laboratory and George Mason University, and by the Sierra Leone Ministry of Health and Sanitation.
PCR-based analysis.
Two PCR-based methods were used in this study: the FilmArray Global Fever Panel (GFP, BioFire Defense, Salt Lake City, UT, USA) and the Malaria Multiplex Sample Ready assay (MMSR, BioGX, Birmingham, AL, USA). The same volume (200 µL) of blood specimen was used in both systems and from both specimen types.
GFP is a fully automated, nested PCR assay capable of automated extraction of nucleic acids from a blood sample and rapid (one hour) detection of nineteen targets using the FilmArray platform, including P. falciparum, P. vivax/P. ovale, and Plasmodium spp. [15]. Briefly, within two hours of sample collection, 200 µL of capillary or venous blood was mixed with 1 mL of GFP sample buffer. The entire diluted sample (total volume of 1.2 mL) was loaded into GFP pouches, and analyzed on the BioFire Film Array 2.0 instrument according to the manufacturer’s instructions. Detection and identification of the target is made automatically based on the melting temperature of the obtained amplicon. The validity of the run is determined by the system based on the amplification results of controls included in the assay pouch. The GFP has three different assays for Plasmodium: one species-level assay that detects P. falciparum, one species-level assay that detects both P. vivax and/or P. ovale, and one genus-level assay (Plasmodium spp.) that detects all Plasmodium species known to cause malaria in humans. The limits of detection (LOD) for GFP are 180, 150, and 240 genomic copies per mL blood for P. falciparum, P. vivax, and P. ovale respectively (Bio Fire, unpublished data). The FilmArray system software calculates semi-quantitative crossing point values (the fractional PCR cycle number when fluorescence of the sample exceeds the background fluorescence threshold). The manufacturer of the instrument refers to these values as “Cp values.” The instrument’s software does not provide these values to the user but they can be accessed by manufacturer and were used in this study to help interpret the data. Since Cp values and more commonly used term “Ct values” are different name of the same value the term “Ct values” is used for both amplification systems for the purposes of the discussion in this paper. Due to the differences in PCR reaction chemistries between GFP and MMSR assays (two-step nested PCR vs. one-step TaqMan PCR, respectively), the values obtained using GFP assay are generally lower for the same analyte concentrations and cannot be directly compared between these systems.
The MMSR assay is a room temperature-stabilized TaqMan-based real-time PCR assay [16, 17], designed to detect P. falciparum, P. vivax, Plasmodium spp., and RNaseP (sample extraction control) in a single assay. While this assay was not specifically designed to detect P. ovale and other less common malaria species, these parasites may be identified in samples testing positive for the genus specific marker (Plasmodium spp.) and negative for both species specific markers (P. falciparum and P. vivax) [18]. Within two hours of collection, DNA was extracted from 200 µL capillary or venous blood using QIAamp DNA Mini Kit (Qiagen, Germantown, MD, USA); the final volume of extracted DNA was the same as the initial sample volume – 200 µL. Five microliters of the extracted DNA was added to each MMSR tube, previously rehydrated with 5 µL water. Tubes were then subjected to a thermal cycling program as follows: initial incubation at 95°C for 2 minutes; 45 cycles of denaturation at 95°C for 10 seconds and annealing/elongation at 59°C for 1 minute. Fluorescence levels were measured at the end of each cycle. The samples were run using 8-tube strips and each run contained a negative (no template) and a positive (P. falciparum DNA) control. The run results were considered valid when all controls gave expected results. A sample was considered positive for a particular target in the MMSR assay if a sigmoidal amplification curve with Ct value < 40 was observed. The reported LODs for P. falciparum in the genus-specific assay (Plasmodium spp.) were 244-390 parasites per mL DNA solution, depending on whether the lyophilized or “wet” format was used [17]. The reported LOD in the P. falciparum-specific assay was similar in the lyophilized test (244 copies/mL), but was about 10-fold higher in the “wet” assay. The LOD for P. vivax was not reported for the lyophilized, “sample ready” format, but was previously determined for the “wet” assay as 127 parasites/mL [16].
Data from capillary or venous samples with negative RNaseP results (MMSR assay) or invalid FilmArray results were not included in the comparison. However, if the matched partner of the invalid sample showed valid results, data from the matched partner were included in statistical analyses of the population as a whole.
Sample populations returning valid results were compared using McNemar’s chi-square analysis of (self-)paired samples, corrected for continuity. Ct values for positive (matched) samples were compared using paired t-tests and for unmatched populations using unpaired t-tests, assuming unequal variances.